Fuzzy Logics for Multiple Choice Question Answering

We have recently witnessed how solutions based on neural-inspired architectures are the most popular in terms of Multiple-Choice Question Answering. However, solutions of this kind are difficult to interpret, require many resources for training, and present obstacles to transferring learning. In this work, we move away from this mainstream to explore new methods based on fuzzy logic that can cope with these problems. The results that can be obtained are in line with those of the neural cutting solutions, but with advantages such as their ease of interpretation, the low cost concerning the resources needed for training as well as the possibility of transferring the knowledge acquired in a much more straightforward and more intuitive way